Juvenile arrests for drug offences grew 120% between 1992 and 1996, translating to over 200,000 youth in 1996 alone. As a result, the Juvenile System (JJS) is understandably eager to implement effective policies, sanctions, and treatments to break the delinquency-drug connection. Thus, there is a great need for reliable/valid instruments that measure alcohol and other drug (AOD) involvement and the multiple co-occurring problems (e.g., mental health) that ultimately affect intervention efforts by JJS and behavior health treatment providers. While a number of instruments exist, the development of adolescent measures has outpaced the field's ability critically evaluate resultant tools, making the selection more difficult than ever before. Although tools have undergone psychometric testing in treatment populations, none have undergone rigorous testing in JJS. This creates unique problems in JJS, where the assessments and judgements must be as objective and impartial as possible. Standardization is part of that effort. Indeed, there is a serious gap in our knowledge as to how to best assess AOD-involved adolescents in JJS and which tools would be most applicable One toll that promise is the Comprehensive Adolescent Severity Inventory (CASI). The CASI is composed of independent modules each incorporating objective, focused, and concrete questions to assess ten areas of adolescent functioning. Given JJS's interest in the CASI and its impressive psychometric properties within an adolescent treatment population, we believe that the CASI has the potential to meet the needs of JJS (if found to be reliable and valid within JJS). Consequently, this application is submitted to examine the following within juvenile arrestees: 1) pre-sentencing, post- sentencing and pre/post-sentencing reliability; 2) concurrent validity of the Alcohol/Drug, Legal and Mental Health Modules; 3) predictive validity; and 4) the empirical refinement of the CASI's scoring system. We also propose to: 5) complete analyses on our current treatment sample as we are now completing follow-up interviews; and 6) conduct generality analysis on the current treatment sample and proposed JJS sample to assess how reliability and validity are enhanced or maximized in these two settings. Three hundred twenty juvenile arrestees will be entered into the five year study over a 38-month period. Male and female youth will be a racially diverse group and will participate in two CASI interviews pre- and post-sequencing, a one- and six-month follow-up and a DISC-IV interview. Urine screens will also be collected. Cronbach's alpha statistic, oblique multiple-group principal-components cluster analysis, factor analysis, multiple discriminant analysis, canonical variance analysis, kappa statistic, and intraclass correlation coefficients will be the primary statistics used.